Our inspiration...

Tennis shoes. A canned beverage. A lava lamp. Something called your attention, and you want one "exactly like that". In hopes to find what you saw, you type in details of the item in your search engine. You spot an image... but it doesn't tell you the brand or where to buy it.

... and our solution for you!

With Sendis, we aim to assess two problems. Our application captures an image of the wanted item using the phone's camera, and then proceeds to find it online. Then, it asks how many do you want delivered straight to your home and does so. We save you all the fuzzy work when you tell us: "Please, send this!". Also, as search is generally prompted by words, Sendis breaks the language barrier that can difficult this process.

Our technologies...

We built the back-end using Python's OpenCV library and Flask, and implemented Firebase authentication. We created Machine Learning algorithms and fed them enough instances of certain items to enable the program to accurately recognize them when a user submits a picture of those. The front-end was created purely in React Native. Finally, we used the NCR API to organize and display the items in the catalogue, and GitHub to collaborate throughout the development of the product.

Our challenges...

Many of the things we did were firsts for us, which made the entire development process very testing. However, our greatest challenge was bringing together the different platforms we used for our project, as being first-time users of those made us uncertain on whether things would work as we expected them to. We take pride in facing this challenge with perseverance and patience, since we proved we can accomplish difficult or unfamiliar tasks due to our determination to do so.

and the lessons they taught us...

The takeaways from this weekend were both technical and teamwork-related. We learned how to use APIs we hadn't dealt with before, how to connect the various platforms and how to adapt a client's prompt to an idea and and use their very own interface. We learned to leverage our teammates' strengths and help them through their weaknesses, and we believe this was the key to how incredibly fulfilling this experience was for each team member. Each one of us discovered areas they were interested to keep learning about, and that makes us eager to keep developing Sendis as further specialized developers. We believe the next step for Sendis is to teach it to recognize more items, and eventually release it to the market.

Share this project:
×

Updates